Background
More than 30,000 people are diagnosed with oral cancer annually in the United States, and due to a poor survival rate that has remained relatively static, of those who are diagnosed, approximately 7,500 die of this disease each year [
1]. Many of these patients will subsequently develop regional tumors, as a result of spreading of the disease to other organs and lymph nodes [
2].
Risk factors are anything that significantly increases the chance of a person developing a disease, such as oral cancer. Scientists now agree that the use of tobacco and alcohol, human papillomavirus (HPV) infection, and in the case of lip cancers, exposure to ultraviolet light, are risk factors that increase the likelihood of developing oral cancers [
3]. Thus, many oral cancers are believed to have a multifactorial etiology in which environmental and genetic cofactors influence the development and progression of the disease. Recent evidence suggests that among these factors, nutrition may play an important role in retarding the development and progression of oral cancers [
4,
5]. A recent meta-analysis examining the association between fruit and vegetable consumption and the incidence of oral cancer, found a nearly 50% reduction in oral cancer risk for each additional portion of fruit or vegetable consumed per day [
6].
It is generally accepted that fruit and vegetable intake may be correlated with a reduction in the incidence of oral cancers even among tobacco and alcohol users. The knowledge gained thus far, however, does not provide enough specific detail upon which to base effective or comprehensive nutrition or dietary intervention recommendations. More detailed information is needed to identify the active components of fruits and vegetables, to evaluate dose-response relationships, and to adjust for potential confounders. The stratification of these studies, for example, has revealed that citrus fruit consumption was associated with a larger protective effect against oral cancer than overall fruit consumption, and was greater than green- or overall vegetable consumption [
6].
Based upon these and other observations, a growing interest has been developing in alternative and complementary medicine to identify botanical and dietary components for their chemopreventive and chemotherapeutic potential. One such class of compounds, the proanthocyanidins (PACs), are polyphenolic compounds that are highly concentrated in certain dietary fruits and nuts, such as grapes, blueberries, cranberries, and almonds, as well as cacao beans and chocolate [
7‐
10].
PACs belong to a larger class of abundant, plant-derived compounds, flavonoids, which appear to provide many beneficial health effects, largely due to their antioxidant properties. PACs have been shown to protect against oxidative stress and tobacco-induced DNA damage, and to exhibit selective cytotoxicity against some human cancers, including breast, lung, prostate, and gastric carcinomas [
11‐
15]. Extracts from grape peels, grape seeds, and black raspberries, sources that contain high concentrations of PAC, have demonstrated selective suppression of tumorigenic phenotypes in oral cancers, specifically in oral squamous cell carcinomas (OSCC) [
16,
17].
The possible mechanisms of PACs effects on tumor proliferation are complex and are likely to involve multiple pathways, which may differ according to the type of cancer. For example, PAC derived from grape seed, green tea, and lowbush blueberry, exhibited broad, anti-proliferative properties against multiple cell lines, including colon [
18], prostate [
14,
19‐
21], breast [
22,
23], and oral cancers [
24]. Grape seed extract (GSE), a rich source of proanthocyanidin, was found to be a potent inhibitor of aromatase activity, an enzyme expressed in higher levels in cancerous than in normal breast tissues [
23]. GSE also induced apoptosis of MCF-7 and MDA-MB-235 breast cancer cells through modulation of both Bcl-2 and MEK/ERK signalling pathways [
22]. GSE and other flavonoid extracts were also found to induce apoptosis in colon, leukemia, and oral cancer cell lines via caspase-3 activation [
15,
18,
24], and were reported to inhibit ornithine decarboxylase (ODC), an enzyme involved in epithelial tumor cell proliferation [
13,
25,
26].
To further evaluate the potential relationship between oral cancer inhibition and PAC, we sought to identify any specific or dose-dependent effects that may exist between the administration of GSE-derived PAC and the proliferation of oral cancers. In this way, we hope to more definitively characterize the effects of this non-toxic agent for its chemopreventive or chemotherapeutic value. Quantitative and dose-dependent relationships between PAC administration and oral cancer proliferation could, therefore, be established prior to the identification and elucidation of the pathways involved, and prior to animal and in vivo studies of this compound.
Based upon the observation that flavonoid-rich extracts, including PAC, inhibited proliferation of multiple tumor cell lines, this study examined the potential for PAC to modulate the proliferative phenotype of OSCC in vitro. While the aforementioned studies have increased our understanding of flavonoid- and nutrition-based inhibition of oral cancer proliferation, none to date have adequately investigated the role of PAC in modulating the proliferative phenotype of OSCC. This study tested the hypothesis that PAC modulates, in a dose-dependent manner, the proliferative phenotype of OSCC in vitro.
We determined that PAC significantly reduced the proliferation of the OSCC cell line, CAL 27, in vitro. Furthermore, this inhibition was dose-dependent and was more pronounced in this oral cancer cell line than in either of the cervical cancer cell lines that we evaluated, Ca Ski and GH354. In addition, the maximal growth inhibitory concentration (GIMAX) of PAC was sufficient to induce a net loss of cells (cytotoxic) in CAL 27, but not in either Ca Ski or GH354 cell lines.
Our previous work established that infection of OSCC with high-risk HPV 16 strongly promoted the proliferative potential of OSCC
in vitro [
27]. Therefore, we further sought to determine if PAC was capable of modulating HPV-driven OSCC proliferation
in vitro. Our results provide one of the first demonstrations of HPV modulating the OSCC proliferative response to PAC. We also found that PAC inhibited HPV-positive OSCC proliferation and that the inhibitory concentration of PAC, providing the maximal proliferation inhibition and cytotoxicity (GI
MAX), was the same for both HPV-positive and HPV-negative OSCC. These results suggest that PAC may exhibit selective inhibition of OSCC proliferation and that PAC administration may be sufficient to inhibit the increased proliferation of OSCC in the presence of HPV 16.
Standard chemotherapies and radiation treatment, frequently used to treat oral cancer patients, often have harsh and severe side effects that include tissue damage and malfunction. The discovery and validation of novel treatments and prevention strategies, lacking these harmful side effects, is one of our main short-term goals. This report provides evidence that PAC may provide one possible complementary or alternative therapy for oral cancer patients, as a safe, effective, cost-saving, and non-toxic treatment. We propose that this study provides compelling in vitro data to justify the exploration and elucidation of the mechanisms of PAC-induced proliferation inhibition of oral cancers and for testing of PAC in animal models.
Methods
Materials
Grape seed proanthocyanidin (PAC) extract (Lot #3717HF7361) was obtained from GNC Preventive Nutrition
® (Pittsburgh, PA). This commercial source of PAC was extracted from U.S. grown wine grapes,
Vitus vinifera. Such commercial sources of grape seed PAC extract are demonstrated to contain 95% PAC and contain approximately 80–90% oligomeric PACs, including dimers, trimers, tetramers, and a small amount of monomeric flavonoids [
28,
29].
Cell lines and cell culture
Human oral squamous cell carcinoma (OSCC) CAL 27, human cervical carcinoma (CC) Ca Ski, human cervical adenocarcinoma (CAC) GH354, and human foreskin fibroblasts (hFF) Hs27 cell lines were obtained from American Type Culture Collection (ATCC: Manassas, VA). CAL 27, GH354, and Hs27 cells were maintained in Dulbecco's Modified Eagle's Medium (DMEM) with 4.0 mM L-glutamine adjusted to contain 3.7 g/L sodium bicarbonate, 4.5 g/L glucose, and 110 mg/L sodium pyruvate, obtained from HyClone (Logan, UT). Ca Ski cells were maintained in RPMI-1640 medium with 2 mM L-glutamine, adjusted to contain 1.5 g/L sodium bicarbonate, 4.5 g/L glucose, 10 mM HEPES, and 1.0 mM sodium pyruvate, obtained from ATCC. Media was supplemented with 1% Penicillin (10,000 units/mL)-Streptomycin (10,000 μg/mL) solution and 10% fetal bovine serum (FBS), obtained from HyClone, except that GH354 was supplemented with 20% FBS. Cells were cultured in 75 cm2 BD Falcon tissue-culture treated flasks (Bedford, MA) at 37°C and 5% CO2 in humidified chambers.
Transfection
CAL 27 cells were seeded in 25 cm
2 BD Falcon tissue-culture treated flasks in appropriate media as described above and allowed to achieve 70% confluence. Cells were then transiently transfected by adding 1 μg/mL of the full-length HPV type 16 virus, cloned into the pBluescript SK-vector (ATCC #45113). The transfection was performed using the Stratagene Mammalian Transfection Kit (La Jolla, CA) according to the manufacturer's recommended protocol for CaPO
4 transfection, and as previously described [
27]. Transfected CAL 27 cells (CAL 27-TF16) were maintained in the same cell culture conditions as described above for non-transfected CAL 27 cells. Mock transfectants of CAL 27 (CAL 27-mTF) were also established by performing the transfection protocol, but without using virus.
Microscopy of cell morphology and viability
Cells were visualized with a Zeiss Axiovert 40 inverted microscope (Gottingen, Germany), and images were captured at 100× magnification with a Canon PowerShot G6 digital camera (Tokyo, Japan). Digital images were subsequently processed using Adobe Photoshop (San Jose, CA) Image Analysis tools. In brief, several wells of unstained Ca Ski, GH354, CAL 27, CAL 27-TF16, and Hs27 cells were photographed at each time point of the proliferation assays (day 1 – day 4) and at each concentration of PAC (0 – 100 μg/mL) to visualize any effects on cell morphology. In addition, several wells of cells were also fixed at these time points using 50 μL of 10% buffered formalin and subsequently stained using crystal violet 1% aqueous solution (Fisher Scientific: Fair Lawn, NJ) to quantitatively document cell morphology, percent of cell spreading, and confluence. Additionally, at each time point, several wells were stained using Trypan Blue (Sigma: St. Louis, MO) and live cells were enumerated to determine viability, as previously described [
30,
31].
Proliferation
In vitro proliferation assays of CAL 27, CAL 27-mTF, CAL 27-TF16, Ca Ski, GH354, and Hs27 cells were performed in the appropriate complete media, with and without the addition of PAC. The total concentrations of PAC used were between 10 and 100 μg/mL. Cells were plated in Corning Costar 96-well assay plates (Corning, NY) at a density of 1.2 × 104 cells per well, maintained at 37°C and 5% CO2 in humidified chambers, and their proliferation was measured over four days. Cultured cells were fixed every 24 hrs (day 1 – day 4) using 50 μL of 10% buffered formalin, and were stained using crystal violet 1% aqueous solution. The relative absorbance was measured at 630 nm using a Bio-Tek ELx808 microplate reader (Winooski, VT). Data were analyzed and graphed using Microsoft Excel (Redmond, WA). Three separate, independent replications of each experiment were performed.
Statistics
The differences between treatments were measured using a
t distribution, α = .05. All samples were analyzed using two-tailed
t tests as departure from normality can make more of a difference in a one-tailed than in a two-tailed
t test. As long as the sample size is even moderate (>20) for each group, quite severe departures from normality make little practical difference in the conclusions reached from these analyses [
32].
RT-PCR
RNA was isolated from 1.5 × 107 cells of CAL 27, CAL 27-mTF, and CAL 27-TF16, using ABgene Total RNA Isolation Reagent (Epsom, Surrey, UK) and the procedure recommended by the manufacture. RT-PCR was performed with the ABgene Reverse-iT One-Step RT-PCR Kit (ReadyMix Version) and a Mastercycler gradient thermocycler (Eppendorf: Hamburg, Germany) using the following primers synthesized by SeqWright (Houston, TX): HPV16 forward primer, ATGTTTCAGGACCCACAGGA; HPV16 reverse primer, CCTCACGTCGCAGTAACTGT. One μg of template RNA was used for each reaction. The reverse transcription step ran for 30 min at 47°C, followed by denaturation for 2 min at 94°C. Thirty-five amplification cycles were run, consisting of 20 sec denaturation at 94°C, 30 sec of annealing at 58°C, and 6.5 min of extension at 72°C. Final extension was run for 5 min at 72°C. Reaction products were separated by gel electrophoresis using Reliant 4% agarose gels (Cambrex: Rockland, ME). Bands were visualized by UV illumination of ethidium-bromide-stained gels and captured using a Kodak Gel Logic 100 Imaging System and 1D Image Analysis Software (Eastman Kodak: Rochester, NY).
DNA isolation
DNA was isolated from 1.5 × 107 CAL 27 and CAL 27-TF16 cells, cultured with and without PAC, with the GenomicPrep DNA isolation kit (Amersham Biosciences: Buckinghamshire, UK), using the procedure recommended by the manufacturer. DNA was separated by gel electrophoresis using Cambrex Reliant 1% agarose gels and the bands were visualized by UV illumination of ethidium-bromide-stained gels and captured using a Kodak Gel Logic 100 Imaging System and 1D Image Analysis Software.
Discussion
Proanthocyanidins are part of a larger group of compounds that have been demonstrated to have anti-proliferative and anti-tumor effects [
13,
26,
33]. PAC can be isolated from multiple sources and the therapeutic benefits of PAC may be determined, in part, by the fruit or vegetable from which it is derived [
34]. GSE and other fruit-derived flavonoid extracts have demonstrated robust anti-proliferative effects on other oral and epithelial cancers [
17,
24,
35,
36].
Although other studies have demonstrated that particular fractions of GSE or raspberry extracts may be cytotoxic against oral tumor cell lines, none have yet examined PAC specifically to analyze its effects on oral cancer proliferation. Through this study, we have demonstrated that PAC exhibited dose-dependent inhibitory and cytotoxic effects on cervical cancer cell lines, with these effects more selective and intensely specific for OSCC. In addition, HPV, now believed to initiate oral cancers among non-smokers, and to modulate OSCC proliferation and progression among smokers, was modulated and reduced, in a dose-dependent manner, by the administration of PAC.
A closer examination of PAC administration among these various cell lines revealed that the effects of PAC were less intense and more gradual in the foreskin and cervical cancer cell lines, Hs27 and Ca Ski, respectively. In contrast, at even the lowest dosages, PAC substantially inhibited the proliferation of both the cervical cancer cell line, GH354, and the OSCC cell line, CAL 27. Furthermore, the level of inhibition was dissimilar between these two cancers – only limiting proliferation of the cervical cell line, GH354, but resulting in a net loss of cells in the oral cancer line, CAL 27. This level of cytotoxicity may represent a selective effect which suggests that OSCC may be rendered more susceptible to this treatment.
Although we found that PAC exhibited a general inhibitory effect on all cell lines tested, a morphological change was observed after PAC administration in both the cervical and oral cancer cell lines that suggested the possible induction of apoptosis. Images obtained from live cell cultures under PAC administration demonstrated that many of the cells were clustering, rounding, and blebbing – morphologic changes that are often suggestive of apoptosis. However, once the cells were fixed and stained to quantify these effects on proliferation, only Ca Ski cells and non-cancerous controls Hs27 were sufficiently adhered to remain in the assay plates at the GIMAX concentration of PAC. One possible reason for this phenomenon may be the degree of progression toward apoptosis or cell death; the more susceptible cell lines, such as GH354, CAL 27, and CAL 27-TF16, could be further along the apoptotic pathway and therefore less capable of adhering and more likely to be removed by mechanical forces during the fixation and staining process. The DNA analysis revealed fragmentation, but no banding patterns indicative of internucleosomal cleavage, suggesting further experimentation will be necessary to delineate the pathways and mechanisms associated with these changes.
Aside from demonstrating that PAC may exhibit differential effects among different cancers, we found convincing evidence that HPV may modulate, and in turn be modulated by, the effects of PAC with respect to the OSCC proliferative response. Our previous research found that HPV 16 strongly induced OSCC proliferation and our findings here indicate that HPV 16 also influenced OSCC response under PAC administration. The introduction of HPV 16 into these cells not only increased their relative levels of proliferation, but also mediated their response to PAC, with lower dosages actually stimulating proliferation of CAL 27-TF16 cells. Interestingly, the GIMAX was determined to be invariable for both transfected and non-transfected cells, suggesting that despite this modulating effect, the HPV 16-specific effect was not sufficient to alter OSCC proliferation in response to incremental increases of PAC administration at or beyond GIMAX.
One critical goal of cancer treatment is to find chemopreventive and therapeutic agents that selectively target cancer cells without detrimental effects to healthy cells and tissues. Although many chemotherapeutic agents are currently being used in the treatment of oral cancers, such as cisplatin (CDDP) and 5-fluorouracil (5-FU), it is known that most tumors may be more or less resistant to these drug therapies because they are not dependent on a single receptor or signal transduction pathway for growth and progression [
5]. Thus, compounds or combinations of compounds that may activate multiple apoptotic pathways, such as PACs, may provide new and more effective treatment regimens for patients with oral cancer.
Recent studies have determined that PACs can be administered orally as a dietary supplement and are subsequently bioavailable, in both serum and tissues at μg/mL concentrations, without any significant toxicity, implying that PACs may represent one class of promising candidates for use as an adjuvant therapeutic agent in patients receiving chemotherapy for oral cancer [
37,
38]. The supporting evidence that GSE and other flavonoid extracts are effective as adjuvants and treatments for breast, colon, and prostate cancer, combined with the results of this study, demonstrate that PAC may act even more selectively on OSCC than other cell lines. This suggests the possibility that PACs may act synergistically to induce selective inhibition and apoptosis of oral cancers
in vivo.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
KK and SO conceived, monitored and coordinated the experimental design. SO, MK, KC, DF, DJ, JP and AS carried out the transfection, cell culture, and proliferation assays. Both KK and SO contributed equally to the writing of this manuscript. All authors have read and approved the final manuscript.